I noticed that people sometimes report R² (‘R-squared’) instead of the Pearson correlation coefficient r when discussing the correlation between two variables. In the special case of a simple linear relationship this numerical equality is not strictly wrong, yet presenting R² as if it were the correlation coefficient might wrongly give the impression they are the same thing. In this post, we will therefore unpack the difference between these two measures, explain their mathematical definitions and proper usage discuss the best practices for when to use each in statistical reporting.